3b G. V- HAMILTON 



ject's knowledge. It represents the outer aspect of an exit 

 door, set within its frame, and equipped with a button (B, Fig. 2) 

 and two strings (S-i, S-2, Fig. 2). When the experimenter 

 pulls the string attached to the inner end of the button (S-i, 

 Fig. 2) the latter is brought to the vertical position and the 

 door is thus unlocked. When the string attached to the outer 

 end of the button (S-2, Fig. 2) is pulled, the button assumes 

 the horizontal position, thus locking the door. The subject has 

 nothing to do with the manipulation of these strings, which are 

 carried in metal eyes to the under surface of the apparatus, 

 whence they are carried in grooves, thoroughly concealed, to 

 any part of the laboratory from which the experimenter can 

 operate them without the subject's knowledge. Each of the 

 four exit doors is thus equipped with a button and two strings. 



The sides of the apparatus are of wire netting. The inner 

 surfaces of the exit doors are painted greyish-white, whilst the 

 remainder of the interior of the apparatus is dark green. The 

 exit doors have no individually distinctive marks. 



The dimensions of the apparatus are determined according to 

 the following general rule: the subject's length being " i," the 

 uniform inside height of the apparatus is "3;" the height of the 

 entrance and exit doors is " 2.9 " and their width " 2." From 

 point O (Fig. i ) to the mid-point of each exit door's lower margin 

 is " 6." The various doors are sand-papered to prevent binding. 



In order to meet special conditions I have made various un- 

 important modifications of the apparatus just described. The 

 human subjects, by reason of their upright position in walking, 

 required a relatively higher apparatus. For the horse an appa- 

 ratus was provided which had " exit doors " merely wide enough 

 to enable him to thrust his head and neck into a food box just 

 beyond; after a trial he was led out of the entrance end of the 

 apparatus. The monkeys so resented handling that their appa- 

 ratus had to be built within a larger enclosure which, in turn,' 

 adjoined their living rooms. In every case the following essen- 

 tial condition was met: the subject, once within the apparatus, 

 was unable to discover whether a given exit door was loclced or 

 unlocked except by actually pushing against it. 



Description of method, (i) Preliminary training. It was found 

 to be desirable to give the animal subjects thorough familiarity 

 with the apparatus before their formal trials were begun, hence. 



